Facebook has updated its resource explaining how it categorizes content on its website, including Feed, Stories and Reels.
Meta has also announced new tools and features to improve transparency on the social network, which will better help marketers.
Why we care Marketers and advertisers need a good understanding of how Facebook ranks content so they can make informed decisions about their campaigns to achieve maximum visibility and engagement.
What’s new? Meta released a number of new features designed to provide greater clarity about its ranking factors through its digital writing:
System Cards: Facebook has created 14 system cards to help marketers understand how Facebook uses artificial intelligence to classify content and create personalized feeds. The cards also explain how people can control what they see.
“Why am I watching this?”: Meta will be expanding this feature to Facebook Reels in the coming weeks. It allows people to understand how their previous activity on the site has influenced the content that AI finds relevant to them and then posts to their accounts.
“Show more, show less”: Facebook plans to make this feature, which is currently available on all Feeds, Videos and Reels posts via the three-dot menu, more prominent.
Content library and Meta APAnd: Facebook plans to release a new set of tools for researchers called the Content Library and Meta API in the coming weeks. The new library includes data from public posts, pages, groups and events on the social networking site.
System cards
Facebook’s new System Cards are the biggest update to its Resource Center. This system consists of 14 cards:
Feed: Facebook uses AI to calculate a relevance score for about 500 posts and then ranks them in descending order. The system is designed to display a variety of content in the feed, meaning a user should not see multiple video posts in a row.
Feed Classified Comments: The AI ranks the comments in order of what it thinks will be most relevant to each user. It does this by examining factors such as the popularity of other comments and whether they have been posted by someone in your network.
Feed recommendations: AI will determine what content users interact with the most based on factors such as groups they’ve recently joined and posts they’ve liked. It then uses this information to decide what content (eg posts, reels, live videos) to recommend.
Rollers: Artificial intelligence selects which reels are posted and in which order by determining what the user is most likely to be interested in. It makes these predictions by examining factors such as which accounts the user has followed, liked, or recently interacted with.
Stories: The AI system automatically shows the stories of people or pages by predicting what the user is likely to be interested in. The system also enforces rules to ensure users receive a balanced mix of content in Stories.
people you may know: The AI tried to determine people who might be of interest by looking at factors such as people who are friends of a user’s friends or people who are in the same groups as the user.
Video: When users watch and interact with Facebook Video, one of the underlying AI systems offers a variety of video types that can match their preferences. This content is located in the Video tab. It may include reels, music, games or shows. This is content users might be interested in from creators they might not follow.
market: When a user views and interacts with Facebook, including the Facebook Marketplace feed, one of the underlying AI systems recommends relevant Marketplace listings. For example, users can see items for sale in categories such as home goods, pet goods, and sporting goods. Users’ feeds may also include other recommendations, such as sellers and content they may be interested in.
Notifications: The AI chooses which notifications to send and ranks them in order of what it thinks will be most relevant to the user. Meanwhile, previously viewed notifications are displayed in the order they were received.
Search: The AI gives each potential search result a score related to the relevance of the content to a user by examining factors such as content type. It will then show users results in order of relevance based on that score
Groups feed: AI automatically determines which posts appear in the group feed and in what order, scoring content for relevance.
Individual group feed: Artificial intelligence predicts what content users interact with the most and then ranks it by relevance in their individual group feed. Relevance factors include what and who users have recently followed, liked or interacted with.
Suggested group: Facebook’s AI will analyze factors such as the groups a user’s friends are members of and topics related to products a user has recently engaged with, and then use that data to identify other groups that may be of interest
Pages You May Like: AI will suggest pages to follow based on pages a user’s friends have recently liked or pages that may be related to products and posts the user has recently interacted with.
Get the daily search newsletter marketers trust.
Content library and Meta API
Another big update to the Facebook Resource Center is the new Meta API and Content Library. The database is configured to include data from:
Public publications Pages Groups Events
People will be able to use the library to search, explore, and filter in a graphical user interface or through a programmatic API.
However, in accordance with current guidelines, this tool has been created specifically for researchers from qualified academic and research institutions working on scientific research or public interest topics. To access this data, researchers will have to request it.
Personalization of the user experience
Facebook confirmed that in addition to providing greater transparency into its ranking factors, it also wanted to give users the tools to take back control of the content they see, such as the “Why am I watching this?” feature.
These tools give Facebook users the ability to shape their own experiences and choose what they do and don’t want to see. People can make changes by visiting their Channel Preferences on Facebook and also through Settings.
What did Facebook say? Nick Clegg, Meta’s President of Global Affairs, shared details about the Meta digital writing about how AI classifies content and how it will be easier for users to control what they see moving forward. He said:
“[Our AI] The systems make it more likely that the posts you see will be relevant and interesting to you. We’re also making it clearer how you can better control what you see in our apps, as well as testing new controls and making others more accessible. And we give more detailed information to experts so they can better understand and analyze our systems.” “Our AI systems predict how valuable content might be to you, so we can show it to you sooner. For example, sharing a post is often an indicator that you found that post interesting, so predicting that you will share a post is a factor our systems take into account.” “As you can imagine, no single prediction is a perfect indicator of whether a post is valuable to you. So we use a wide variety of combined predictions to get as close as possible to the right content, including some based on behavior and some based on user feedback received through surveys.” “We hope to introduce these products to early development researchers. process, we can get constructive feedback to make sure we’re building the best possible tools to meet their needs.”
Deep dive: You can find a more detailed explanation of the AI behind the content recommendations on the page AI Meta Blog. To learn more about how AI uses signals to make predictions, you can visit Meta Transparency Center.
[ad_2]
Source link
